Q&A with Jesse Gomez: How the brain learns to recognize faces


A set of experiments in kids overturns conventional wisdom, and finds that parts of the brain grow as we learn to recognize faces.

By Stacey Huang

Jesse Gomez is a sixth year graduate student in the lab of Kalanit Grill-Spector, a professor of psychology. He studies development of the areas of the brain involved in recognizing faces and reading—parts of the visual cortex. Traditional neuroscience wisdom holds that a child’s brain prunes itself over time as the child learns. Gomez and his colleagues recently challenged this conventional wisdom.

Gomez discussed these results and also how he uses Pokémon to study brain function with Stacey Huang, a graduate student in electrical engineering.

What did you do?

The main question we were asking was: “How does your brain change as you get better at something?” So, we started collecting data on kids between 5 and 12 years old. We put them into an MRI machine to measure the amount of brain tissue and activity in the visual cortex as they looked at visual stimuli like faces and words. We also scanned graduate students as a control.

What did you find?

The conventional wisdom in neuroscience is that you're born with more brain tissue than you need, and it's pruned away over time. Surprisingly, we saw that there was less tissue in the face region of the kids’ brains compared to the adults’, which means tissue has to grow from childhood to adulthood. My first thought was, “Oh no, did I just do something wrong?” But we looked back at the original study from 1986 about pruning and realized it had only examined a very limited part of the brain.

We also found that more tissue in the visual cortex correlated with a better reading and face recognition ability. That suggested that tissue growth is related to brain function.

By “tissue growth”, do you mean that the region in the brain gets bigger?

The region isn’t getting bigger; rather, there is growth within the region. Imagine the brain is like a garden and the flowers and plants are like brain cells. The garden itself is not growing in acreage, but the things in it are getting more complex—they're growing more branches and leaves.

You were able to make this discovery by studying kids, who are often overlooked in neuroscience studies. What are some of the challenges in studying kids?

We have to train them. The first time they come in, we put them in a fake MRI machine where they have a magnet on their head, and if they move, then the Disney movie they're watching stops. So they learn what “staying still” means. There is also a test we use in adults to map brain activity in the visual cortex. People have tried to do the test in kids but haven't succeeded because it's too boring. I actually turned that into a video game for the kids, where they have to help this alien fly home by filling its ship with gas. This kept their attention long enough for us to complete the experiment in kids for the first time, which is really cool.

You’re now doing a study on how the brains of Pokémon experts light up when they see Pokémon. Why Pokémon?

It sounds like a silly experiment, but it answers a really fundamental question for why the brain is organized the way it is. Everyone has the same brain folds in the same locations; for example, our face and place recognition regions can always be found on the same brain folds across people. A big question is, “Why do these regions always appear in the same place?” Scientists suspected it had to do with something inherent about the visual stimuli, either the size or the curviness of the object, but didn’t have a good way to test these competing theories.

I realized that people in the 90’s played the Gameboy Pokémon games a lot and thus their brains might specialize in recognizing Pokémon. Pokémon are perfect visual stimuli; they’re smaller than faces and places, but in curviness they fall somewhere in the middle. So if a brain region specializes for Pokémon, then the area that lights up should pop up between the faces and places folds if curviness is the key factor and outside them if size is the key factor.

What I found consistently across the experts is that there's a Pokémon-selective region that pops up outside the face and place regions. This fits the size hypothesis—so the visual cortex is organized based on how big the images on your retina are.

What got you interested in studying development of the visual cortex?

No one tells you to look at someone's face. You just learn on your own, and most of us get really good at distinguishing people’s faces. We’re not good with all faces though. Take monkey faces for example—you glance at a group of monkeys and think they all look identical. But monkeys are experts at monkey faces. This suggests that what you experience is sculpting your brain, and I’m really interested in how that happens, and how the brain is wired to do this automatically. It's also interesting to ask: “What happens when stuff goes awry? What happens in development in people who aren't good at faces or reading?” Those are questions I'm looking at too.